From: Pete Herzog (lists@isecom.org)
Date: Wed Feb 22 2006 - 05:00:00 EST
Hi,
> fingerprints) are not encrypted so their content can be easly sniffed
Easily perhaps from many internal networks. But it's much
more difficult for an attacker to sniff it without access to
either the client's network and the server's network. It's
not impossible but difficult enough to warrant OpenBSD pf
attention as a type of protection.
> out. Port knocking increases sealthness (so it can probably be
> considered a security measure) but cannot be used to make a system safe.
Stealthiness is privacy. Privacy is a loss control and does
make an operation safer. The efficacy of the privacy loss
control, as there can be degrees like as in encryption,
however does take issue. So whether or not this type of port
knocking is good enough is really the question and not whether
it protects at all because it does. Even obfuscation is
protection albeit not a good one but it will minimize some
types of attacks. This is in many ways like Anti Virus or the
IDS that relies on black lists. They narrow the kind of data
which enters a host or network by authorizing that which it
does not know to reject. Would you consider either of those
technologies to not be a form of protection? At least port
knocking is white list :)
> I've noted that many people misunderstand the port-knocking's meaning,
> it's not a security layer and doesn't prevent security holes, it can
> only make your system quite harder to hack.
But it is a security layer because it makes a system harder to
hack. How is that not a security layer?
> Anyway the use of port-knocking to make a system safer can be a bad
> solution, because it's just like wrapping a vault with cellophane. You
> can use it to hide a computer in internet, or to keep your logfiles
> clean as you can use cellophane to protect iron against rain.
Well then it does protect the vault from rain, right? It's
still protecting. Not every loss control is going to be of
the bunker type. Again, look at any black list technology and
you'll see the same thing. I have done years of narrow
studies on this and I can prove that simply moving a service
to a different port will provide an every day layer of
protection. But not every attack. Then again, it's not
supposed to. It's only supposed to fend off the automated
which is already a lot. This is why internet crime is so much
easier than say, cat burglary because of automation they can
find many more potential targets.
>
>
> Gimeshell, it's not right to consider tcp options size unlimited, and I
> think the real port-knoking is harder to brutefoce. Authentication
> through os fingerprints cannot use all tcp options (p0f doesn't analyze
> them all) and, anyway, it must fit tcp options structure and respect the
> protocol rules.
This is true but it doesn't dismiss the concept as a layer of
protection. It's also not an industrial solution and it
doesn't scale well at all. But for some purposes, it's
possible to implement and maintain inexpensively.
Sincerely,
-pete.
------------------------------------------------------------------------------
Audit your website security with Acunetix Web Vulnerability Scanner:
Hackers are concentrating their efforts on attacking applications on your
website. Up to 75% of cyber attacks are launched on shopping carts, forms,
login pages, dynamic content etc. Firewalls, SSL and locked-down servers are
futile against web application hacking. Check your website for vulnerabilities
to SQL injection, Cross site scripting and other web attacks before hackers do!
Download Trial at:
http://www.securityfocus.com/sponsor/pen-test_050831
-------------------------------------------------------------------------------
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:55:33 EDT